Markets and Households on Low Incomes

Transcription

1 Markets and Households on Low Incomes Europe Economics and New Policy Institute September 2010 OFT1268

2 Crown copyright 2010 This publication (excluding the OFT logo) may be reproduced free of charge in any format or medium provided that it is reproduced accurately and not used in a misleading context. The material must be acknowledged as crown copyright and the title of the publication specified.

3 CONTENTS Chapter/Annexe Page 1 Executive summary 4 2 Introduction 11 3 The low income population 14 4 Food 21 5 Energy 28 6 Financial services 42 7 Transport 70 8 Internet 77 9 'Enabling' products Conclusions on markets and low income households 91 Annexe A: List of interviews 102 The work was undertaken by Europe Economics and New Policy Institute for the Office of Fair Trading. The views expressed in the publication are those of the authors and not necessarily those of the Office of Fair Trading.

4 1 EXECUTIVE SUMMARY Introduction 1.1 This report provides a review of how people on the lowest incomes are treated in a number of case study markets and identifies reasons for any disadvantages that they suffer compared with people on higher incomes. We considered a range of factors on both the demand and supply sides of these markets which might contribute to any disadvantage. 1.2 The report also considers whether people in low income groups have less access to certain 'enabling' products, such as bank accounts and the internet, which provide improved access to other products and whether they are, as a result, disadvantaged in other markets. 1.3 We have principally relied on official data sources and published material. This has been supplemented by interviews with experts in the relevant sectors. Low income households 1.4 The low income population is defined here as people living in households where the household income is in the lowest 20 per cent of all household incomes. In practice, this is very close to the official measure of income poverty (that is an income less than 60 per cent of median household income). Identifying the bottom 20 per cent as 'lower income' in this way allows us a common approach across different data sources. 1.5 The likelihood of an individual belonging to a low income household varies significantly between groups. Groups where this likelihood is substantially above the average include workless working age households, lone parent households and those living in social rented accommodation. Children and pensioners also face an above average likelihood (or 'risk') of belonging to a low income household. By contrast, groups where this likelihood is not particularly pronounced can nevertheless make up a substantial 'share' of the overall low income population. Working-age adults and those living in major urban areas are two examples in this category. This distinction, OFT1268 4

5 between 'high risk' groups that are over-represented in the bottom fifth of the income distribution and 'high share' groups which account for a significant proportion of that bottom fifth, runs through this report. Market reviews 1.6 We have reviewed five main goods or service markets food, energy, financial services, transport and internet access. This provides coverage both of essential goods and of 'enabling' products. Food 1.7 In the food market evidence on whether people on low income pay more is inconclusive. Local convenience stores are more expensive or may provide lower quality than supermarkets largely reflecting the economics of operation and buying power of the large chains. There may be some localised market power. Low income households without cars (a notably large share of whom are lone parents) are less able to access the special offers and lower prices available in large supermarkets. Cultural aspects can benefit low income households with better local provision of food in some ethnic communities. Energy 1.8 In the energy market price and choice rather than access or quality are the main issues. Vulnerable groups, including those on low income are not accessing the cheapest tariffs to the same extent as other groups. Prepayment is the most expensive tariff and is used by a quarter of low income households. The standard tariff which is also widely used is only slightly cheaper. The lowest tariff is offered for payment by direct debit with further reductions for online billing. Ofgem considers that the prepayment tariffs are broadly cost reflective but that standard tariff charges are not cost reflective. Licence conditions have been introduced to require cost reflectivity. Doorstep selling by energy companies has led to significant numbers of customers switching to more expensive tariffs. Lack of a bank current account and of internet access, both of which are more likely for low income households, make it difficult to access the cheapest deals. OFT1268 5

6 Financial services 1.9 We have reviewed a number of financial services products. The percentage of low income households that do not have a bank current account is almost double the national average. Basic bank accounts have been developed but do not have the full range of features available with a current account. The accounts on offer may not meet the budgeting and transaction needs of people on low income. For those with current accounts insufficient funds charges levied by banks bear heavily on low income and vulnerable customers. Product revenues and costs are misaligned and there is cross subsidy from those paying charges to people on higher income A range of factors can inhibit the opening of accounts risk of incurring high charges, self- exclusion, lack of financial literacy, mistrust of financial institutions and onerous ID requirements can all play a part Low income households, particularly those living in social housing, have a lower than average take up of home contents insurance with many saying that it is not affordable. Premiums both for home and car insurance, based on risk of claims, tends to be higher in low income areas. Insurance products may not be well suited to the needs of people on low incomes and there is some distrust of the providers Low income households without a bank account or with low credit rating may be limited to high cost forms of credit. This market is well developed but for some forms there is only a limited range of suppliers. Provision of small loans, with clear repayment terms and without credit checks meets the needs of this group, although at high cost. Individuals tend to stay with the same provider and forms of credit with which they are familiar rather than shopping around. Lack of a bank account, lack of confidence and difficulty in understanding the charges involved all act to deter switching We also looked at the extent to which low income households had any savings which might help them avoid the need to take out high cost loans. Nearly half of all low income households have no savings. These households may also rely more heavily on informal savings mechanisms, including cash OFT1268 6

7 and savings stamps. Easier to understand savings products could encourage greater use. Transport 1.14 Car ownership is seen as another enabling product allowing access to wider employment opportunities and better access to cheaper food outlets. Over half of people in the lower income group do not own a car compared with about 25 per cent in the population as a whole. The marginal cost of travelling by public transport is higher than that of private transport. People on low income may also be deterred from travelling because of the risk of incurring extra cost if public transport is cancelled or conditions of cheap fares are not met. Internet 1.15 Access to the internet has increased rapidly for all income groups over the past decade but only a third of low income households have access compared with 60 per cent of the population as a whole. Take up is lowest amongst the elderly and households in social housing. Evidence suggests that the reasons for not having internet access are no longer financial but more due to lack of confidence, skills or motivation. Substantial cost savings are potentially available through use of the internet but having home access does not guarantee that those will be achieved. Enabling products 1.16 Holding a bank current account, car ownership and internet access provide benefits in their own right but also enable the user to get benefits in other markets. This is noted above in the findings from individual markets. It is difficult to quantify this benefit but evidence on food expenditure suggests that car ownership may allow worthwhile savings for some low income households. Low income households with internet access spend a smaller proportion of their income on energy. OFT1268 7

8 Conclusions on markets and low income households 1.17 We have found a number of ways in which low income households can be disadvantaged. This can be in terms of price paid, quality of product and ease of access. Three groups are most likely to be disadvantaged across a range of markets. These are lone parent households, people living in social rented housing and people with disabilities On the supply side of the markets we have identified a range of factors which can explain why the lower income group is disadvantaged. These include: market structure which allows the exercise of market power in a number of the markets reviewed a variety of pricing practices including cost related and risk based pricing which vary between markets. These are consistent with competitive market behaviour but may still work to disadvantage people on low income suppliers decisions on location, for example, on out of town supermarkets, which may be driven by commercial considerations but still result in disadvantage for this group of customers product design, particularly in financial services may not meet the needs of the low income group. It has taken government intervention to encourage the development of a basic bank account innovations have occurred in a number of the markets reviewed which benefit low income households. This can be seen in local food initiatives, new, albeit high cost, forms of credit and online products Innovation and product design should be stimulated in a competitive market but it is difficult to judge whether more would have occurred if there had been more competition. An alternative explanation could be that limited demand for such products meant that they were not commercially viable. The case for government intervention to stimulate new products is then more about equity than market structure. Internet development is a major OFT1268 8

9 continuing source of market led innovation but increased access is necessary if low income households are to benefit fully from this On the demand side there is much more commonality of explanatory factors indicating that people in the lower income group face similar problems and have similar responses in the markets reviewed Not surprisingly the budget constraint is a strong influence on this group's behaviour. More surprisingly loyalty to traditional suppliers deters switching supplier despite paying higher charges. This may in part be because the higher cost products better meet the needs of this group. It may also be related to product costs and other terms and conditions not being easily understood. Energy and financial services involve more complex and less frequent decisions than weekly food shopping. The provision of good information about the choices available is correspondingly more important. That is not an issue confined to low income households but the impact of over-payment will be more significant for that group Concern about penalties that might be incurred by going into debt combined with low savings lead to higher risk aversion in the lower income group. This can lead to low take up products such as current accounts with direct debit facilities which could bring wider benefits. Instead they opt for high cost products such as prepayment meters and extended warranties that carry lower risk of facing unexpected charges. With low income there is a premium attached to having certainty about regular payments There are also other behavioural features which can mean that low income households get less good value. Mistrust of institutions and lack of confidence feature in the financial services and internet markets. Cultural, social or educational issues have been identified as affecting consumer behaviour in the food, energy and financial services markets. Loyalty to traditional suppliers is also in part a behavioural trait. These are not all negative. Provision of ethnic minority foods can improve availability for low income groups The demand side factor which runs across all of the markets considered is the disadvantage which comes from a significant proportion of poorer OFT1268 9

10 customers not taking advantage of (or being able to take advantage of) the wider benefits which flow from use of key 'enabling' products. Lack of access to products in one market can result in adverse price effects in another market While difficult to quantify, the potential for cross market improvements in the position of poorer customers through improved access to 'enabling' products may have greater impact over a period of years than intervention on the supply side in individual markets. Improving the interactions between markets in order to achieve a better functioning of markets as a whole, may provide OFT with new policy options which go beyond its traditional focus on issues within individual sectors. OFT

11 2 INTRODUCTION 2.1 This report has been commissioned by the Office of Fair Trading (OFT) in order to provide an overview of how people in the lowest income group are treated in specific markets and to identify reasons for any disadvantages that they suffer compared with people in higher income groups. 2.2 The OFT has no legal or other mandate to redistribute wealth but one of the factors that it can take into consideration is whether an intervention would benefit disadvantaged customers. This study will therefore provide the OFT with background and context to help it in setting priorities for future work. 2.3 In carrying out the study we have looked in detail at definitions of poverty and how the numbers of households with low income vary according to demographic factors. We have focused on households in the bottom fifth of the income distribution, which we generally refer to as the lower income group, providing a relative rather than an absolute measure of poverty. We have then reviewed how this lower income group fare in their involvement in individual markets and how factors on the supply and demand side of each market affect that outcome. 2.4 People with lower incomes will, as consumers, generally have less choice than people with higher incomes. Not only do they face a tighter budget constraint with a higher proportion of their expenditure devoted to the essentials of food, clothing, housing and energy but they are also less likely to have savings to cover unexpected expenditure and are also less likely to have access to key products such as bank accounts, personal transport and the internet which act as 'enabling' products by improving access to and choice of products in other markets. 2.5 Analysis of how the lower income households are served by particular markets therefore needs to consider the extent to which people in the lower income group, as a whole or in part, are disadvantaged by receiving a less good 'deal' in terms, for example, of the price and/or quality of particular goods and services than people in higher income groups. OFT

12 2.6 For the most part we have drawn on published material but we have also carried out interviews with experts in each of the markets covered in order to test our findings and obtain additional material. Where and why are people on low income at a disadvantage? 2.7 This report starts with a review of statistics on low income households and particular demographic characteristics of the lower income group. We then reviewed five main goods or service markets food, energy, a range of financial services, transport and internet access. Food and energy were selected because they account for a significant proportion of the expenditure of low income households and both markets have been the subject of considerable research into the treatment of people on low incomes. Financial services and access to the internet have been identified by government as markets which help to foster social inclusion. They also, along with transport, play a wider enabling role in improving the access that people on low incomes have to other markets. 2.8 Between them these case studies illustrate the variety of ways in which the lower income group can be disadvantaged and provide the basis for more general conclusions. There are many reasons why people in the lower income group may be disadvantaged and these are likely to vary between markets. During the course of this study we have identified a number of factors on the supply and demand side of markets which are relevant to the outcome. These are listed in generic terms in Table 2.1. In the final sections of the report we set out a more detailed consideration of the relevance of these factors both within and between markets and look, in particular, at the role of the enabling products. OFT

13 Table 2.1: Supply and demand side factors affecting the lower income group Supply side Demand side Market structure Budget constrained price/quality trade off Cost of supply Mobility Price discrimination Limited information Pricing linked to payment method Risk aversion Risk based pricing Lack of savings Location Lack of confidence/mistrust of institutions Terms and conditions Inertia/constraints on switching Marketing practices Cultural/social/educational Product design Lower access to enabling products Innovation Product requirements OFT

14 3 THE LOW INCOME POPULATION The scale of poverty 3.1 This section presents an overview of the low income population the 'poor consumers' in the UK relative to the population as a whole. All of the data and definitions in this section come from the Department for Work and Pensions' Households Below Average Income (HBAI) series. 3.2 In this project, we identify the low income population with people and households in the bottom fifth of the income distribution. This identification of the poorest fifth with the low income population is justified below. These income distributions is based on household incomes after income tax and national insurance, but before housing costs have been deducted. 3.3 Using income before housing costs (BHC) does carry certain problems, the greatest of which is how to deal with Housing Benefit. In a BHC measure, Housing Benefit is treated as income, when in reality it is hypothecated. Moreover, a household would see its BHC income increase if its rent, covered by Housing Benefit, were to increase. An After Housing Costs (AHC) measure does not suffer from these difficulties as it deducts the rent being paid from the BHC figure. 3.4 However, it is the BHC measure that the Government is using in the Child Poverty Bill, currently going through parliament. The BHC measure will therefore be enshrined in law as the principle measure of low income. For this reason, we have used the BHC measure of low income in this report. 3.5 In doing so, we note that some groups figure more prominently in a description of low income BHC than AHC. Those with low housing costs (pensioners, people in Northern Ireland) tend to have higher relative risks of low income BHC. For those with high housing costs (some private renters and people with mortgages, people in London) the opposite is true. 3.6 Household income is then adjusted for size and composition. This process (known as equivalisation, and carried out by Department for Work and Pensions (DWP) according to internationally recognised processes) OFT

15 recognises that larger households require more resources, but that the relationship between size and need is not exactly linear. While a household with two people may need twice as much food as a single person household, it does not need twice as many cookers or refrigerators, for instance. 3.7 Table 3.1 gives values for different points along the BHC income distribution for different household types. Table 3.1: Values of quintile points of the income distribution for different family types ( s per week 2007/08) Household type Equivalisation factor 20% point 40% point Median 60% of median (low income threshold) Couple no children Single adult no children Single adult two children aged 5 and Couple, two children aged 5 and Source: Households Below Average Income, 2007/ The figures in the table are unequivalised cash values based on the income distribution of the entire population. So for example, a couple with no children is in the bottom fifth if their weekly income is below 244. By contrast, a single adult with two children is in the bottom fifth if its weekly income is less than A couple without children is used as the benchmark household type, and given an equivalisation factor of1. OFT

16 3.9 One important point of note is that that the values for household income at the bottom quintile are very close (within a few pounds per week) of the values for 60 per cent of median income for respective household types. 60 per cent of the median is the conventional low income/income poverty line used by the government and the proximity of this value to that quintile point for the bottom fifth is what justifies our approximation of the bottom fifth with those in low income For simplicity, then, the analysis that follows uses households in the bottom fifth as its definition of low income. This is simpler to understand than the 60 per cent of median definition, but quantitatively little different. Moreover, whilst the HBAI dataset would allow us to use the 60 per cent definition, other datasets simply divide the income distribution into fifths. Looking at the bottom 20 per cent thus allows us to make comparisons within and across different data sources. Features of low income households 3.11 We have considered two principal approaches to analysing the relationship between demographic factors and poverty. We distinguish between: 'High risk' groups that is, those who are disproportionately represented in the bottom fifth of the income spectrum including workless households, lone parent households, households with at least one disabled adult and Pakistani and Bangladeshi households. There are very few simple cuts of the population whereby over half of any group is in the bottom fifth. The exceptions are workless households, and Pakistani or Bangladeshi households. 'High share' groups that is groups which account for a significant proportion of the total number in the bottom fifth of the income spectrum. There are many examples of instances where a group has a relatively low risk of low income but, due to its size, makes up a large share of the low income group. White British households are one such group the risk of low income is lower than average, but this group makes up 80 per cent of all those in the bottom fifth simply because it makes up 85 per cent of the total population. OFT

17 3.12 Table 3.2 summarises the findings under these two headings. As above, the data for this analysis are taken from the HBAI Survey published by ONS which is derived from the Family Resources Survey (FRS). Table 3.2: Percentage of the population in low income households Factor High risk groups High share groups Age groups Pensioners (25%) children (25%) Working age adults (50%) Work status Family type Disability Tenure Unemployed (70%), other working age economically inactive (51%) Lone parent (39%), Female Pensioner (32%) Households with a disabled adult (26%) Local Authority rental (43%) Housing Association rental (35%) Head or spouse aged over 60 (25%), Other Working age economically inactive (25%), Unemployed (10%) Working age couple with children (32%) Households with no disabled adult (66%) Owned with a mortgage (23%), owned outright (32%) Ethnicity Bangladeshi, Pakistani (both 50%+) White British (81%) Rurality Major Urban areas (22%) Major Urban areas (38%) Source: Households Below Average Income, 2007/ Particular features to note are: Age groups: - Half of those in the lowest fifth are working age adults and a further quarter are children per cent of children are in the bottom fifth of the income distribution, as are 25 per cent of pensioners. - The figures for working age adults with and without children are 20 per cent and 15 per cent respectively. Work status: - Inevitably, the 'risk' of being in the bottom fifth of the income distribution is much higher among people living in workless than working households. OFT

18 - Seventy per cent of people in households where the head of the household is unemployed and 51 per cent of those in other workless working age households are in the bottom fifth of the income distribution. - These high proportions are quite rare, in that they are among a handful of examples of more than half of a particular group being in the bottom fifth of the income distribution. - Between them, unemployed and other workless households account for around one third of all people in the bottom fifth of the income distribution. Households where the head or spouse is retired account for a further quarter. - This means, though, that a significant minority of those in the bottom fifth of the income distribution live in a households where at least one adult is in paid work. Tenure: - Around two-fifths of people living in social rented accommodation (either local authority or housing association accommodation) are in the bottom fifth of the income distribution. - This compares to around 10 per cent of those who own with a mortgage and 23 per cent of those who own their homes outright. - Those who own their own properties outright make up almost one-third of the low income population. Those who own with a mortgage make up a further quarter. The overall distribution 3.14 Having looked at the low income population, we now look at the overall population broken up into income deciles. This allows us to look at both OFT

19 those who are very poor the bottom 10 per cent, rather than the bottom 20 and those who are nearly poor For some demographic features the income distribution is heavily skewed. This is the case for people in workless working age households, lone parent households and occupants of local authority and housing association accommodation. The distributions for each group are shown in Figure 3.1 below. Figure 3.1: Income distribution of different household types Source: Households Below Average Income, 2007/ It was noted above that over half of people in workless, working age households were in the bottom fifth. In fact, almost all of this group are in the bottom half of the income distribution. This is also broadly true of lone parents and their children and people in social rented accommodation It is not true, though, of pensioners. Despite being over represented in the lowest two deciles (around one in eight of pensioners are in each of the OFT

20 bottom two deciles), around one quarter are in the top four tenths of the income distribution It is worth noting, though, that people in workless households are most heavily represented in the bottom decile, whereas the other groups are more prevalent in the second bottom decile. As such, we see that people in workless households are more likely to be very poor In fact, adults in workless households without children are almost by definition in the bottom tenth of the income distribution. If they have no other incomes, benefit levels for these groups are low enough to guarantee that they will be in the bottom 10 percent Pensioners are in fact slightly more likely to be in the third lowest decile than either of the two bottom deciles. That is to say, there are slightly more 'nearly-poor' pensioners than 'very poor' pensioners. OFT

21 4 FOOD Main points Issues concerning the poor paying more for food are controversial and evidence on differential prices is not conclusive. Local convenience stores are more expensive than supermarkets, reflecting economics of operation and buyer power of larger outlets. There has been a movement of supermarkets away from city centres where building sites are less expensive. An access problem for the poor reflects an interlinkage between food and transport markets, as a car acts as an enabler for access to supermarkets. 'Food deserts' in the traditional sense may not be applicable today but issues relating to the shopping experience persist. There may be areas with elements of local monopoly power (only one convenience store within walking distance), but this is very location specific. The culture of food, seeing food shopping as a social experience, is important. This comes through from observed differences in the quality of food outlets in low income predominantly ethnic areas versus low income non-ethnic areas. Low income lone parents without cars spend significantly more on food than those with cars. Second degree price discrimination in the form of 'bulk buys' is a likely explanation, as price per unit of food comes cheapest in bulk buys, which is particularly difficult for lone parents without a car to carry and/or store. OFT

22 Introduction 4.1 It is important to recognise that the issue of 'whether the poor pay more for food' is controversial and not universally agreed upon by researchers. Alternative hypotheses have been met with reservations and complications due to lack of broad empirical data, the complexity that arises in trying to measure nutritional intakes and changing patterns of the large retailers. There is also a distinction between research based on 'mapping' an area, and research using first hand expenditure data. Access, price paid and quality 4.2 Food poverty, as defined by The East Midlands Health Observatory is: If a household or individual are unable to obtain a nutritionally adequate diet, they are typically considered to be living in food poverty There is evidence to suggest that those without a vehicle, and not within walking distance of a large grocery store, face higher food costs. According to a study by the Food Standards Agency (Scotland, ): 3 There is a tendency for prices to be lower in larger shops and in areas with a low level of social and economic deprivation. 4.4 We were told that prices in the smaller high street branches of the major supermarkets are typically five to 10 per cent higher than their larger counterparts. 4 But these stores may still be less expensive than the local convenience stores. 4.5 The Scottish study found a negative correlation between the availability of healthy foods and level of deprivation (as deprivation increases, number of available healthy foods falls). This study showed that the price for a healthy foods basket varied across store type, median price ranging from in 2 Flaherty (2009) 'understanding food poverty'. 3 Food Standards Agency Scotland (2008) 'Accessing healthy food: A sentinel mapping of healthy food retailing in Scotland'. 4 Interview. OFT

23 large stores, in medium sized stores to in small stores. The area with the highest prices was rural, at per basket. The basket could be purchased cheapest in the two affluent small town areas. 4.6 UK households in the lowest income decile tend to spend the highest proportion of income on food: 26 per cent versus 15 per cent for the rest of the population but in absolute terms, these households spend much less. 5 An earlier study found that the poorest fifth spend 25 per cent of their total income on food, and in order to purchase a healthy diet would need to spend over 30 per cent A recent US study suggested that poor households pay less than richer households for identical goods. 7 The study found that the poor pay less because they shop in cheaper stores and because they pay less for goods in the same store, as they are more likely to buy food on sale. The poor shop more in convenience stores, where prices are higher, but have a higher share of expenditure in supercenters where prices are lower than in grocery stores. The findings of this study run counter to the conventional comparative shopping basket analysis. Further research would be necessary to see whether similar results can be observed in the UK. 4.8 In the UK, during the late 1990s, areas characterised by deprivation, social exclusion, and poor access to healthy, affordable food became known as 'food deserts.' A 2003 study found that after a Tesco was built in the area of Seacroft and Whinmoor of Leeds, which is in the top five per cent most deprived wards in England, those living within walking distance significantly increased their fruit and vegetable consumption A 2007 study suggested that more extensive empirical investigations of food deserts in the UK have found very little evidence suggesting that areas 5 Dowler (1997) 'Budgeting for food on a low income in the UK: the case of lone parent families'. 6 Leather (1996 ) 'The making of modern malnutrition' Caroline Walker Lecture. 7 Broda, Leibtag, Weinstein (2009) 'The role of prices in measuring the poor's living standard' Journal of Economic Perspectives. 8 Wrigley, Warm and Margetts (2003) 'Deprivation, diet, and food-retail access: findings from the Leeds 'food deserts' study' Environment and Planning A 2003, volume 35, pages OFT

24 exist with large proportions of residents with poor access to retail food stores Quality is an issue in the sense that the UK poor often receive less than optimal nutrition because of access and affordability. In 2008 the Department for Environment Food and Rural Affairs (DEFRA) reported lower percentages of fruits and vegetable consumption in poor households, with two adult households in the lowest three income deciles purchasing about 650 grams per person per week (about one fruit per person per day) less than households on middle income. 10 The Scotland study revealed that in terms of quality, ratings on fresh vegetables and fruits are better for large general stores, while small stores and stores in deprived areas have the greatest proportion of items rated as poor quality Other surveys in the UK found that few low income groups have difficulty accessing and affording fruit and vegetables. Few also claimed that they had trouble accessing supermarkets. 11 Reasons for differential outcome Supply side 4.12 One interesting aspect of the food market is that suppliers clearly compete for customers at the lower income end Over the past 20 years the large supermarket chains have accounted for an increasing share of sales. There has been a growth in large, often out-oftown stores. Specialist local grocery shops have declined in number while there may have been a small increase in small convenience stores. That 9 Macintyre (2007) 'Deprivation amplification revisited' International Journal of Behavioural Nutrition and Physical Activity. 10 DEFRA (2008) 'UK purchases and expenditure on food and drink and derived energy and nutrient intakes in 2007'. 11 Dibsdall, Lambert, Bobbin, Frewer (2003) 'Low-income consumers' attitudes and behaviour towards access, availability and motivation to eat fruit and vegetables' Public Health Nutrition 12 This was raised in one of the interviews in reference to numerous budget lines and special offers aimed at those less willing to spend. OFT

25 includes the development of local outlets by the large supermarket chains. The development of larger outlets away from residential areas has increased travel times with increased reliance on car transport In rural areas with no relatively affordable grocery store within walking distance, people without a car are significantly more excluded than owners of a car in the same neighbourhood. In urban areas where public transport is available, there is often a problem for the poor as they face time and income constraints and/or they have children which makes public transport difficult Without access to a car, it is also difficult to access food at lower marginal costs (in special offers and 'bulk buys' which are a form of second degree price discrimination). According to one of our interviewees, it is the middle class that benefit from, and are the target market for, bulk buy stores such as Aldi or Lidl because they are more likely than the poor to have transportation to carry and store large packages. Those without a car are less able to benefit from the lower marginal costs, and face higher transportation costs, as they require more grocery trips, compared to those able to buy large quantities at a time. Lone parents without a car are especially disadvantaged as it is more difficult for them to carry children and large packages on public transport or on foot Low-income areas may be less attractive to large retailers because of lower average expenditure. According to Leather (1996), the small shops that tend to locate in poorer areas have higher food prices because they have higher operation costs and more market power, along with less availability of items. The lower turnover in shops in lower income areas means it is often not worthwhile to stock a range of fresh fruit and vegetables. Buyer power also comes into the explanation, as covered in the recent Competition Commission investigation. 13 We were told that the buying power of the smaller stores in the wholesale market had diminished substantially to the detriment of prices in the smaller stores. Medium-sized store groups have slightly more buying power but it does not match that of the larger chains. 13 Competition Commission (2008) 'The supply of groceries in the UK market investigation' OFT

26 4.17 According to a study in Leeds, local polarisation in the geography of poverty in the mid 1990's fuelled the problem of the poor facing higher prices, as poor households became increasingly concentrated in small areas and the net incomes of the richest grew much faster than the poorest We were told in interview that in Northern Ireland fewer people have access to cars and the public transport system is much poorer making access to food outlets more difficult. Demand side 4.19 Fresh fruit and vegetables tend to be more expensive than equivalent less healthy options. 15 So that when faced with a budget constraint a low income family may often choose low price over high nutrition in order to feed their entire family. Individuals doing shift work or working non-standard hours may also rely more on convenience food Cultural factors can also be important. In ethnic minority areas (both nonwhite and white such as Italian or Polish), there tend to be a larger selection of outlets selling healthy food at reasonable prices. 16 We were also told that shopping for food should be considered as a social experience this is particularly important for pensioners and can create a community feel. The quality of the outlet itself also becomes important in this regard. Recent changes 4.21 We were told about a growing movement of food co-operatives in lowincome areas. These co-ops supply food to local people, and many are aimed at the poor and charge lower prices. This is not formally part of the market sector, and can be government funded. For example, we were told that the Welsh Assembly Government funded 167 co-ops. There is an 14 Wrigley et al (2003) 'Deprivation, diet, and food retail access: findings from the Leeds 'food deserts' study. Environment and Planning A. 15 An example given in an interview was of a bag of ready to cook vegetables plus other ingredients to make a meal costing more than burger and chips (before accounting for the additional time and labour costs involved in the cooking). 16 This was stressed by more than one interviewee. OFT

27 element of instability with these programmes. If local authorities rely on food co-ops rather than the issue of prices being higher at local convenience stores, and lack of a supermarket, there is a danger that volunteers move on and the co-op dies out, the community is left with no co-op and no supermarket Other initiatives are being undertaken to change this culture of food. Farmers markets, co-ops, community cafes and grow your own initiatives can all contribute to this shift. Local authorities, whether through government or lottery funding, do have the power to make a difference, for example through planning regulations. OFT

28 5 ENERGY Main points Price and choice, rather than an access or quality, are the main issues. Vulnerable groups are not accessing the most competitive tariffs to the same extent as non-vulnerable groups. Rises in fuel prices in recent years have led to an increase in fuel poverty. The majority of those in fuel poverty are pensioners. Spending is lower in local authority and RSL property than owner occupied or privately rented accommodation, in part reflecting the public sector commitment to improving energy efficiency. Prepayment is the most expensive form of payment, being much higher than direct debit, but not differing greatly from standard credit. The proportion of low income households using prepayment is around one quarter, which is double the average for households as a whole. Note also that the majority of the fuel poor use standard credit. Standard credit customers as well as those in social group E are the least active in switching. Not all the difference between prepayment meters is cost-reflective, although Ofgem believe they are broadly reflective. Standard credit charges are not cost reflective. Doorstep selling has led to a significant proportion of low income customers switching to worse tariffs. Cross effects with other markets lack of internet access and bank accounts can restrict access to lower cost options. OFT

29 Price paid, choice and fuel poverty 5.1 In the energy market the issue for low income households is not about access to electricity and gas but about the terms on which energy is available and the absolute size of the household fuel bill. Ofgem's energy supply probe found that vulnerable groups, which include those on low incomes, those without a bank account and those without easy internet access, were not accessing the most competitive tariffs to the same extent as non-vulnerable groups. 5.2 A household is said to be in fuel poverty if it needs to spend more than 10 per cent of its household income on fuel to heat the home to an adequate level of warmth (where adequate is usually taken to mean 21 degrees Celsius in the living room area and 18 degrees in other occupied rooms). The level of fuel poverty will be affected both by changes in income levels and by changes in fuel prices. Fuel poverty will also be affected, over a longer timescale, by any improvements in the energy efficiency of housing. Rises in fuel prices in recent years have led to an increase in the number of households in fuel poverty. In England, this has risen from 1.2 million in 2003 to 2.8 million in 2007, or from 5.9 per cent of households to 13.2 per cent. 17 Estimated required energy expenditure 5.3 Table 5.1 shows the average required spend on heating and light for different income quintiles for This year is used here because it is the most recent for which such a breakdown by income can be provided. 17 Department of Energy and Climate Change (2009) 'Annual Report on Fuel Poverty Statistics 2009' OFT

30 5.4 The estimated average annual fuel bill to cover home heating, water heating, cooking and lighting varies a little between income groups from 990 for the bottom income quintile to 1,132 for the top quintile. 18 Table 5.1: Average required spend on heating and light by income quintiles Group Average required annual fuel spend Whole population 1,029 Bottom income quintile nd rd 1,004 4 th 1,063 Top income quintile 1,132 Source: English House Conditions Survey, These differences largely reflect house size and type of occupancy. These estimates can be broken down by family type, economic status, tenure and location as shown in Table Required spend is highest in couple households, and lowest in single person households. It is also higher among working than non working households. This is simply because working households tend to live in larger houses. The reason the gap is not larger is that workless households are assumed to spend more time at home and heat their homes during this time. 5.7 It is notable that spend in local authority and RSL (Registered Social Landlord) property is much lower than in owner occupied or private rented accommodation. This is true for the bottom fifth income group as well as the population as a whole. This is likely to reflect the public sector 18 The costs in the table are the estimated costs required to keep the home at a comfortable level of warmth. As such, they are modelled figures, not actual expenditure. They take into account the energy efficiency and size of the home, as well as the method of payment. For more information, see Updates and Modifications to the Fuel Poverty Methodology for the 2007 Fuel Poverty Analysis, OFT

31 commitment to improving the energy efficiency of the housing stock which has not been matched in the private rented sector. 19 But it also reflects the fact that social accommodation is relatively smaller. 5.8 All else equal, the cost of heating private rented accommodation should be no higher than for social renters, yet it is some 20 per cent higher in some cases. Here, the interaction of the housing and energy markets, with lower energy efficiency in private rented accommodation, results in a worse outcome for some low income consumers. 19 The English House Conditions Survey for 2006 indicates that around nine per cent of LA housing has a low energy efficiency rating, compared to around 15 per cent of private rented accommodation. OFT

32 Table 5.2: Required fuel spend by households with different characteristics Group Average required fuel spend per household Average required fuel spend for households in bottom quintile. Total 1, Couple with dependent child(ren) 1,199 1,113 Couple, no dependent child(ren) 1,068 1,039 Lone parent with dependent child(ren) 986 1,004 One person aged 60 or over One person under Other multi-person household 1,075 1,064 1 or more work full time 1,074 1,095 1 or more work part time 1,057 1,010 None working and none retired None working, one or more retired Local authority housing Owner occupied 1,088 1,044 Private rented 1,017 1,112 RSL Town and fringe 1,055 1,047 Hamlet & isolated dwellings 1,687 1,364 Village 1,420 1,281 Urban > 10k Source: English House Conditions Survey, 2006 OFT

33 Payment methods 5.9 There are three types of payment for gas and electricity direct debit, standard credit (that is, pay by cash or cheque) or prepayment. The average costs of gas and electricity for a standard quantum of energy paid for by different payment types in 2009 are shown in Table Table 5.3: Prices of gas and electricity by method of payment, 2009 Standard Credit Direct debit Pre payment Gas Electricity Source: Department for Energy and Climate Change Prepayment was found to be the most expensive form of payment, especially for gas, where it was 25 more per annum than standard credit and nearly 100 more than 'offline' direct debit. However, the difference between prepayment and standard credit for electricity is small. The main price differential is with direct debit A notable price differential has also been reported between 'online' and 'offline' direct debit with the former exhibiting cheaper tariffs. With an online direct debit scheme, customers opt to receive and pay bills online and sometimes to supply their own meter reading. It is estimated that some 1.3 million customers, or five per cent of households in Great Britain, made use of this scheme in Ofgem estimated this online discount is around 50, with a range of between 60 and 150 between 2004 and Ofgem's evidence points to a cost differential between the two methods in the region of 10, suggesting strategic motivation behind the heavy discounting. Indeed, Ofgem observed that the online tariffs of the big six 20 The 'average' is based on consumption of 18,000 kwh gas and 3,300 kwh of electricity. These figures should not be confused with those used in Table 4.1 and 4.2 which relate to a specific level of comfort. 21 Department of Energy and Climate Change (2009) 'Quarterly Energy Prices September 2009' Tables and Ofgem (2008) 'Energy supply probe initial findings report', Paragraph Ofgem (2008) 'Energy supply probe initial findings report', Paragraph OFT

34 suppliers only remained open to a new customer for a short period of time. A difference of 46 per cent was reported between existing and new customers of online direct debit deals for British Gas' first Click Energy online tariff until October Table 5.4 looks at the proportion of households using prepayment for gas and electricity, by income quintile. The figure for electricity covers all homes. The figure for gas covers only those homes with a gas connection. Table 5.4: Proportion of households using pre payment meters for gas and electricity Group Proportion using pre payment for gas Proportion using pre payment for electricity Whole population 12% 14% Bottom income quintile 24% 27% 2 nd 16% 19% 3 rd 11% 14% 4 th 5% 7% Top income quintile 2% 3% Source: English House Condition Survey, Around one quarter of low income households use prepayment to pay for their gas and electricity. This proportion is double the average for households as a whole The annual report on fuel poverty statistics (2009) 25 commented that standard credit had become the most common method of payment among the fuel poor for both gas and electricity having overtaken prepayment. As is shown in Table 4.3 standard credit was almost as expensive as pre payment. 24 Ofgem (2008) 'Energy supply probe initial findings report', Paragraph Department of Energy and Climate Change (2009) 'Annual Report on Fuel Poverty Statistics 2009'. OFT